In 2025, much of what we see online may not be real. That includes followers, likes, comments, and views.

According to Juniper Research, more than $114 billion in digital ad spend is expected to be lost this year to fake engagement. Bots and AI agents now drive over half of all web traffic, a growing share of which impersonates humans across platforms like TikTok, Reddit, YouTube, and X.

Engagement, once the internet’s core currency, is collapsing under the weight of synthetic participation.

A Crisis of Credibility

In just one quarter, TikTok removed 185 million fake accounts. Facebook purged over a billion. Analysts estimate that more than 60 percent of X’s users could be bots, according to Internet 2.0’s 2024 study. And across social media, AI-generated comments, fake reviews, deepfakes, and synthetic influencers flood timelines and recommendation engines.

Users are noticing. Surveys show growing “content fatigue” and erosion of trust. A 2025 report from CMSWire found that 81 percent of users no longer trust social media content, and over half skip past influencer posts, treating them as noise. Meanwhile, small businesses spend as much as 30 percent of their ad budgets on fake traffic (Tapper.ai).

The implications are both financial and cultural. “It’s not just that ad dollars are being wasted,” says a digital marketing analyst who asked not to be named. “It’s that the entire system of online credibility is breaking down. The metrics we use to determine success are now suspect.”

AI Has Accelerated the Problem

The rise of large language models has dramatically scaled the problem. In 2025, Anthropic’s Claude was used to generate a coordinated network of fake personas on Facebook and X, engaging with tens of thousands of users in what researchers described as “influence-as-a-service.”

Reddit recently banned university researchers who had deployed bots with distinct human personas, including one posing as a trauma counselor, that earned over 10,000 karma points across months of posting. Most users didn’t realize they were interacting with AI.

These are not isolated incidents. A 2025 report from the Institute for Strategic Dialogue found that automated bot networks were responsible for nearly all the shares of certain deepfake political videos in Germany. Platforms, overwhelmed by volume and complexity, struggle to respond in real time.

Gaming Is Becoming the Testing Ground

While much of the internet grapples with how to restore authenticity, the gaming sector may be quietly charting a path forward.

Unlike social media or content platforms, games are structured, closed-loop environments with defined inputs and measurable outcomes. That makes them unusually well-suited to detect and deter fake participation. And as financial incentives in gaming rise, especially in competitive and Web3 titles, so does the motivation to defend them.

Game developers have turned to increasingly aggressive tactics. Riot’s Vanguard anti-cheat operates at the kernel level of users’ systems. Moderation communities on Discord, Reddit, and Telegram enforce code-of-conduct norms. Studios are investing in real-time detection, reputation systems, and behavioral analytics to distinguish bots from players.

In March, gaming hardware firm Razer introduced “Razer ID verified by World ID,” a system that links player accounts to Worldcoin’s biometric identity infrastructure, designed to verify that players are uniquely human. The move sparked debate, especially around privacy and accessibility, but it underscores the urgency with which some companies are approaching the issue.

One Experimental Approach

Among the newer entrants is Coliseum, a Web3-native tournament platform that already boasts over three million community members, with more than 30,000 participating in its app beta. Coliseum integrates verification into the competition layer itself. Rather than depending on biometric scans or client-side anti-cheat software, Coliseum uses a network of human validators and protocol-level logic, a system it calls the Guardian Network, to confirm that participants and outcomes are real.

Players build reputation through verified play. ‘Guardians’ are incentivized to detect manipulation. Only verified actions are rewarded, and wallets accrue credibility that can’t be faked.

“What we’re seeing isn’t just ad fraud or hyper-botted views or engagement, it’s the computational collapse of digital trust,” states Josh Gier, Chief Marketing Officer at Coliseum. “AI has weaponized engagement, making every ‘like’ and ‘view’ suspect. Our approach isn’t about better detection, it’s about a paradigm shift to a ‘proof-of-participation’ economy. We’re leveraging game theory and distributed validation made possible via blockchain technology to make verified human interaction the most valuable asset, creating a blueprint for the next era of digital advertising and community.”

While still early, Coliseum represents a growing school of thought: that fake engagement is not a moderation issue, but an infrastructure issue. If the system makes faking more costly than playing fair, the logic goes, trust can scale again.

The Fight for Digital Trust

The crisis of engagement is no longer hypothetical. In 2025, regulators have started stepping in. The U.S. FTC recently approved rules penalizing the purchase of fake followers or social proof. European agencies have moved against biometric ID systems like Worldcoin for privacy violations. Major platforms now conduct frequent “bot purges” to reassure users and advertisers.

But many believe technical defenses alone won’t solve the problem. What’s needed is a shift in design, platforms that prioritize verified participation by default, not exception.

Gaming may not fix the internet. But it might be where the repair begins.